|
--- |
|
base_model: moro01525/T5_FineTuning |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: T5_FineTuning |
|
results: [] |
|
--- |
|
|
|
# T5_FineTuning |
|
|
|
This model is a fine-tuned version of [google-t5/t5-small](https://huggingface.co/google-t5/t5-small) |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.8659 |
|
|
|
## Model description |
|
|
|
The model is specialized on Text2Text Generation, in particular the model receives an input like "Ingredients: ingredient1, ingredient2, ..." (containing a list of ingredients) and generates a recipe |
|
|
|
## Training and evaluation data |
|
|
|
This model is trained using [**these**](https://www.kaggle.com/datasets/shuyangli94/food-com-recipes-and-user-interactions) datasets |
|
|
|
## Training procedure |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 2e-05 |
|
- train_batch_size: 4 |
|
- eval_batch_size: 8 |
|
- seed: 55 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 1 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:------:|:----:|:---------------:| |
|
| 0.9318 | 0.1818 | 1500 | 0.8757 | |
|
| 0.9498 | 0.3636 | 3000 | 0.8712 | |
|
| 0.9157 | 0.5455 | 4500 | 0.8683 | |
|
| 0.9177 | 0.7273 | 6000 | 0.8672 | |
|
| 0.9295 | 0.9091 | 7500 | 0.8659 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.42.4 |
|
- Pytorch 2.3.1+cu121 |
|
- Datasets 2.20.0 |
|
- Tokenizers 0.19.1 |
|
|